Woo-hoo, we’re approaching the end of this particular project. In this update we present the final* “Player” version of the Expanded Frontier Map, both the color and black and white versions. This version just contains the systems and jump routes known by the UPF as presented in the rules, modules, and some fan content.

(* Final assuming I don’t find any typos and such)

Changes

First up, as I was generating the data for the color version of the map, I realized that there were just way too many neutron stars on the map. There’s one on the original AD map, five on the Zeb’s Guide map, and I had added in a couple dozen or so more.

One is too many for this volume of space. You might find one (anything is possible) but beyond that is really stretching it. However, since the five are in the original materials, I left them in. I also left in two of the ones I added in the lower left of the map. This puts all the neutron stars in the Frontier region and towards the Vast Expanse. Maybe there’s an explanation for that.

For all the other neutron stars, I replaced them with multiple star systems. That changed two dozen stars systems from neutron stars to multiple star systems, some with as many as five stars.

In comments on the previous update, it was pointed out that I forgot to label Mechan, the world in Zeb’s Guide where the Mechanons fled to when they left Volturnus in the Zeb’s Guide timeline. I don’t typically include that event in my games which is why I probably forgot to include it. Mechan is just up and to the right of Scree Fron on the map. That just required changing the label.

In fixing the labeling on Mechan, I also noticed that the Zeb’s Guide map adds another star 3 light years above Mechan near Capella. So I added a star in there as well.

As far as I can tell, that fixes everything.

The Black and White Version

So that gives us the new and updated black and white version of the map.

The Color Version

Creating the color version was mostly a matter a generating a large number of random systems and then positioning them on the map. Since I had to manually update all the positions, that took a couple hours of work to get all the data into the file and the coordinates updated.

After that, I could generate the rough map with my mapping software. The next step was to go through by hand and tweak the position all the labels that needed to be moved. There actually weren’t that many, I just had to move a few names and a few of the jump distances so that they were not overlapping any of the jump route lines. That cleaned up in just a few minutes.

The other major portion of the map was creating all the nebula. Since I had the outlines of them on the black and white map. I just copied (and scaled) those outlines to the color map. In a few places where the B&W map has little wisps of nebula next to the big ones, I connected them as the process I use to create the nebula would make those wisp-like areas.

This actually took less time than I was expecting. Since I was working zoomed out on the map, the rendering went fairly quickly. Whenever I zoomed in, however, it would take some time to redraw the screen due to all the detail in the nebula. From the earlier sample map, which was zoomed in, I had experienced a bunch of slow down and was expecting the same here. I thought I was going to have to do the nebula on a series of layers so that I could turn them off and on. That wasn’t actually necessary and this process only took a couple of hours instead of the several hours I was expecting.

The hardest part for me was picking colors. The truth is that there probably wouldn’t be any nebula in this part of space but since they look cool and were included in the original material, they are part of the map. I tried to pick colors for them based on the stars that were around as most of the nebula would be reflection nebula since none of the stars in the area put out that much ionizing radiation to cause the gas to glow on its own. I’m not completely happy with it but I like it. This is something that might get revisited at some future date.

The next step was adding in a legend. This also might get redone in the future to include each of the different stellar colors but it is good enough for now.

Finally, I decided I didn’t like the flat black background but wanted there to be a bunch of faint “distant” stars speckling the background of the image. To get that effect was pretty simple. I just applied the Textures->Cracked Glass filter to the black background layer. I’ll admit that I got lucky here and it applied the exactl level of background noise that I wanted. There are other filter combinations that would work as well but that was good for now.

With that done, I signed it and exported the final image.

Up Next

The next step is to do the Sathar worlds and jump routes. Since all the stars are there, all I need to do is add in the links and update some labels. That will be the topic of the next and final (at least for a while) update.

Share you questions, comments, and impressions in the comment are below.

Okay. I’ll admit that this project has taken over my brain. It’s been fun for me to work on it and dust off some of my programming skills. This will be another small update as I haven’t had a lot of free time this past week but enough to get some work done. If you haven’t read them yet, the early posts are here: original post, update 1, update 2, & update 3

The good news is that I think there will just be two more updates and we’ll have the completed map. There might be a few more after that for upgrades to the software but those will probably be a bit more spaced out. Here’s what’s left to do:

• Random generation of spectral types (this post)
• Complete “Alpha Dawn” map (this post)
• Complete “Player’s Version” of the Extended Frontier map (update 5) – This is the full color version of the B&W map I’ve been posting
• Complete “Referee’s Version” of the Extended Frontier map (update 6) – This will be a version of the map with all the Sathar system labels and jump routes included.

I’ve decided that to get to the completed map, I’ll just do the nebulae by hand. I already have the shapes drawn on the B&W version of the map so it’s easy enough to transfer them and do the render. Getting the software to do the nebulae will be a future project.

If there is interest, I could also do a remake of the Zebulon’s Guide to Frontier Space version of the Frontier map in this style. While I’ve modified my map to match most of the named systems on the Zeb’s Guide map, they are not exactly the same (for reasons discussed in earlier posts) and the unexplored stars do no match between the two maps. If there is a large clambering for it, I’ll do that as a future project.

Let’s get on with this week’s update:

Random Stellar Types

After I had the program interpolating the colors between the base spectral types, the next step was to sit down and code up the random generation of those spectral sub-types. I needed this to randomly fill in all the “unexplored” systems with a proper mix of spectral types. The explored systems are way too heavily weighted toward the F & G stars and there are a few too many (i.e more than about 3) for the number of stars in the region but that’s okay.

In any case, I wanted the relative numbers to be approximately correct based on real stellar distributions so I busted out data on the stellar mass functions and after a bit of math came up with a d100 table for the various main sequence spectral types (A, B, G, etc.) to determine the relative ratio of their sub-types (A0, A2, A4, etc). Once I had the numbers it was a simple matter to code up a few subroutines to generate the values. There are some refinements I could make in the future but this is good enough for now.

Complete “Alpha Dawn” Map

Once the program was generating random spectral types in the correct distribution, it was time to finish the smaller map corresponding to the Frontier provided in the original printing of the game and the later “Alpha Dawn” reprint (named to distinguish the core rules from the Knight Hawks supplement).

I did this by running a random map and then taking the spectral types generated and applied them to the remaining systems in the original map. Next I took the resulting file and edited it to tweak the positioning of some of the labels so that they weren’t overlapping any of the jump lines or other text. Finally, I took the nebulae I had created for the earlier version and imported them into this map. This gave me my final version of the smaller map.

I might revisit this in the future and do the nebula with different colors but for all intents and purposes, this one is done.

In making this I realized that the stars in system FS18 are really bright. It’s an M0 Giant and A0 main sequence star in a binary system. Combined, they are ~39x brighter than the sun and only 2 ly from Athor. That means they would appear to be at -5.27 magnitude star in the night sky. For reference, the brightest star in our sky, Sirius, is only -1.46 (small, more negative numbers are brighter). Venus, at its brightest, is a -4.2 magnitude object. That means that this star appears 2.5 times brighter than Venus at it’s brightest and about 30 times brighter than Sirius.

The Full Map

Now that the AD map is done, it’s time to start working on the full Extended Frontier map. I knew I was going to need a lot of random star systems so I generated a big random map with the program to get a good number of star systems with the correct spectral type distribution. I’ll use this whenever I need a random system value.

That done, the first step was to get the data for all the various labeled systems on the map. I already had the ones from the AD map I just needed to add the others. The data for these systems came from a variety of places. For the systems included on the Zeb’s Guide map, I used the spectral types listed in Zeb’s Guide after verifying that they made sense when mentioned in the modules (which usually just gave a color). I didn’t find any real issues so I just used the Zeb’s Guide values.

For the Saurian systems (east of the Frontier), the spectral types were taken from the article “The Sauria Sector” published in the Frontier Explorer, issue 4.

The data for the S’sessu systems (south of the Frontier), the spectral types were taken from the article “Phri’sk Anyone? Detailing the S’sessu Homeworlds” in issue 16 of the Frontier Explorer (p 23 – I haven’t gotten around to putting the individual articles of that issue on-line yet). However, that article only listed spectral types for two of the systems. For the other three, I selected random spectral types in the G & K spectral types from the list of random systems.

With all the star systems in place, I added in all the jump routes connecting those systems. I discovered I had a limit in my code that would not print a jump if it was over 14 light years in length. This was preventing the jump between Gruna Garu and Lynchpin from being drawn so I had to fix that bug. I still don’t let the random process create jumps that long but if you’re entering the data for a map and want a longer connection, the software won’t stop you anymore.

This also required me to change a couple of things. If you’re familiar with Zeb’s Guide, you may notice that the connection between Zebulon and Capella is now a straight line at 9 ly instead of a broken line at 10 ly. The broken line never made any sense given the Knight Hawks rules so I’ve done away with it. The other change was the connection between the S’sessu worlds of Kashra’sk and Minan. The map in the Frontier Explorer lists the distance as 12 ly but it is actually only 11. That is fixed on this map.

The next step is to now add in all the other systems on the map. To do this, I simply pull a system entry from the big random list for each star system on the map and remove the system from the big list so I don’t reuse it. I started in on that process, filling in the systems to the southwest of the Frontier but haven’t finished entering everything else yet. That will be the subject of the next update. However, here’s the big map as it currently stands.

This is just the raw, software generated map, with no tweaking or extra annotations on it. Over the next week or two, I’ll start adding in the rest of the stars and the nebulae. This map covers an area 90 ly wide and 100 ly high which should give your players plenty of area to explore if they need to get out of the Frontier and into the wilds.

That’s it for now

That’s where the project currently stands. If you have any thoughts, questions, or suggestions, let me know in the comments below.

It’s been a fairly busy couple of weeks between family vacation and recovering from said vacation. I’ve been working on the map project but a lot of it has been behind the scenes and not anything that is easily displayable although some progress has been made. Additionally, I took the first steps toward getting the Murder at Rosegard adventure actually written up into a coherent document but again, nothing to display. So on to the map work.

Working on the Nebulae

This first bit definitely falls into the non-visible bit. I began implementing code that will allow me to draw the nebula. Over the past couple of weeks I figured out an good algorithm to use to trace the grid boxes containing the nebula and dived a bit into the code for SVG graphics to understand how to draw curvy lines.

There is nothing to see here yet but the next steps that I’ll be working on are:

• Write the code that draws the curvy boundaries for a predefined nebula region
• Write code to read in nebula regions from the data file
• Write code to randomly generated nebula regions from some to-be-specified parameters.

That will probably be the content of the next update.

Stellar colors

The other major bit I worked on was adding in code to interpolate the colors of the stars based on their actual spectral type. Originally, the code could just handle a few spectral types: B0, A0, F0, G0, K0, & M0 plus a few special types like brown & white dwarfs, black holes, and neutron stars. That is what you’ve been seeing in the maps to date.

The colors for those stars were picked partially to be visually accurate and partially to aesthetically pleasing. Basically I had to exaggerate the color variations between the stars as almost all of them would appear white to the human eye. The interpolated colors will smoothly shift between these predefined colors.

What I wanted was to be able to handle any spectral type, i.e. G2, F3, K7, M5, etc. To do this I needed to interpolate the colors down the spectral sequences. It took a bit of reworking the code to do the interpolation and that’s were I spent most of the time I had this week.

Once that was done, I went back through and put in the actual spectral types for all of the named Frontier sector stars. These were derived from a variety of places. In the original rules, the stars were just given colors which correspond roughly to spectral types. Then there were a few more details in the published modules. Zebulon’s Guide to Frontier Space provided actual spectral type designations for all the stars in the Frontier. However, some of those spectral types in Zeb’s Guide didn’t agree with the previously published material.

This isn’t the first time I’ve tried to do this and luckily, the last time I did it, I recorded all the information. That was done on my Star Frontiers Network Wiki in the Frontier Travel Guide section. So for most of the stars, all I had to do was go through the pages on the wiki and grab the spectral types I had already figured out for each star.

With those code updates, and the updated spectral type values, you get the following map of the Frontier.

There are a couple of things missing from this map. First, I didn’t put the nebula back on as this was a test of the spectral colors. Plus I didn’t do anything for Padda or the star systems labeled with the FSnn designations. Those are still at the “0” spectral types. I will get to those but there is another bit of the code I want to update first (more below).

Oh yea, @Fiddleback on Twitter mentioned he couldn’t see his house and I promised him a zoom box.

The Full Map

The other thing I worked on was getting everything set up for the big map of the full extended Frontier region. This mainly involved just translating the systems in this map to the larger canvas.

There’s no difference between this one and the one above, it is just situated on the canvas is the correct position to correspond to the big black and white map. Over the next little while I’ll be adding in all the other systems as I extract their coordinates from the big map into the data file that generates this map. Watch for updates in the future.

What’s Next?

I’ve already mentioned the work on the nebulae that I have on deck. On the spectral type front there are a few things I want to add there as far as functionality goes.

First, (this is the bit I hinted at above) I want to add code to the random system generation to properly generate the 0-9 subtypes. The relative ratios of major types (A, G, M, etc.) are accounted for but I need to get the relative numbers withing the major types correct. I have the data (or can get it) to do that, I just need to crunch the numbers and write the code. Once that’s done, I’ll add real spectral types to the systems labeled with the FSnn designation and really be ready to generate the full map.

With the spectral types the other things I need to do include:

• Add code for drawing non-main sequence stars. Right now it handles all the main sequence stars (luminosity class V) and some giants and super-giants (luminosity classes III and I) but it may need to handle more when I get the full random generator done.
• Add some interpolation of stellar sizes for main sequence K stars. Right now they are set to a fixed size but I’d like the later K stars to be a bit smaller than the G and early K stars and approach the size of the M stars. Again, the data is there in the program, I just need to code the interpolation.
• Fix the color interpolation for M stars. It’s not obvious in the map that there is a problem since I don’t have any M stars that aren’t M0 but as you progress down the spectral sequence I’m interpolating to the colors of the Brown Dwarfs which are a pinkish color (In real life they’d be a really dull red which doesn’t stand out on the map). I want to make the interpolation go to a deeper red instead of pink. This should be a really quick fix. I know what needs to be done, I just need to code it up and test it.

Since I can do the nebula by hand for the Extended Frontier map, this work on the spectral types is where I’ll be focusing my efforts for the next little while. Once those updates are done, I’ll have everything I need to make the full-color Extended Frontier map.

The Code

Again, If you want to play with the code as I’m working on it, you can grab the current version from my StarMapGen repository on GitHub. This is still not the most user-friendly program to run as it it requires coding in all the parameters by hand before running but that’s on my todo list as well.

I had planned to to have the next part of the module campaign write-up as the post this week, but that’s taking me a bit longer than I had planned and I spent more time than I should have on updates to the Expanded Frontier Map so I’m writing about that instead.

As I said in my last State of the Frontier post, one of the things I want to work on this year is a full color version of the Expanded Frontier map. So while I was on vacation, I fired up my laptop and started on that. This post will detail some of the things I’m working on and what to look for in the future.

Corrections

The first thing I had to do was fix some errors I noticed in my original expanded map. Mainly, I had left off some labels on a few of the star systems. I haven’t posted a version of the map with those labels yet but will in the future. So I added those in.

The other omission I had to correct was one I noticed in creating entries for the Detailed Frontier Timeline. I was missing a necessary jump route for the sathar. So I added that in as well. Again, I haven’t posted that version of the map so the correction just appears on my “Referee’s Copy” at the moment.

Next, I decided to shift the system I was labeling as Belnafaer (from the SFAD5: Bugs in the the System module) to the same one that the TSR folks did in the Frontier map in Zeb’s Guide. The choice was arbitrary and both my selection and theirs matched the description given in the module. I figured it was better to disagree as little as possible with the “canon” material. (I’m still not moving Starmist or Rhianna though).

Finally, I discovered that the labeled distance from Madderly’s Star to White Light was wrong. The map labels it as 6 but it is actually 8. (The map in Zeb’s Guide has it correct but the one in the Dramune Run is wrong as well.) This was actually discovered in making the color map as described below. So I fixed that. It’s funny, but I’ve been playing Star Frontiers for 35 years now and never noticed that typo.

Those were the corrections I made. While you can’t see two of them, this is the resulting map:

I also fixed up some internal errors, mostly labels on the wrong layers, but that has no impact on the map itself. I had planned to post the version with all the star designations, but in looking at them, I realized they need some cleaning up on their positioning so it wasn’t quite ready to go for this post. I need to spend about an hour tweaking things.

Moving on to Color

As I mentioned before, I had written a program to randomly generate sector maps. The stars were represented using the symbols in the chart to the right. The second column showing how multiple star systems are represented.

As written, the program only generated random systems and wrote basic map data out to a file. What I needed was the ability to read in that output file and build the map from the data instead of randomly.

So that’s where I started. I had an output format, I just needed code to read that and convert it back into data that could be used for drawing the map. Along the way, I realized that I needed to change some things about the way the data was represented and stored in the program and output file as we well as needing to add storage for things like the name of the system.

Another thing I had to change was the printing of the Z-coordinate. The random map generation process creates star systems in 3 dimensions and labels each system with its Z-coordinate above or below the plane of the map. Since for the Frontier Map (at least initially) everything is on the same plane, there was no need to print little zeros by each star system. So I added a flag to turn off printing that coordinate on the map.

Generating the Data for the Frontier Map

With the code now ready to read data from a file and draw the map. I needed to generate the data for the Frontier map. As I’m building and testing the software, I figured I’d work with a smaller map so I’m just using the original Frontier Map from the Alpha Dawn rules.

There are 51 systems on that map. I determined the coordinates on the map (as the software sees them, currently measured from the upper right corner – that may change), and wrote those into a data file. That was the easy part.

The next step was to get spectral classifications for each of the systems. The inhabited systems were relatively easy. In the listing for each of the systems in the rule book, it gives a color for the star. At this point the program isn’t refined enough to break down the spectral types so a G0 star is represented the same as a G2 or G8 star. So I only had to approximate. Yellow stars got a G0 designation, orange stars got a K0 designation, white stars got an F0 designation, and red stars got a M0 designation. I used those since those are the spectral types that the program currently understands.

Next I went through and added in the spectral types for the stars that host the various planets of the modules and added in designations for them. Then, I had to go in and add in the designations for star systems in the Rim that were labeled in Zebulon’s Guide. Zeb’s Guide give full spectral types (F7, G3, K2, etc.) but for now, I just used the spectral type letter and not the number since the program can’t process those yet. (Note: if you compare the spectral types from Zeb’s Guide for the original Frontier worlds to the colors given in the Alpha Dawn rules, they don’t always agree. That’s something I’ll have to resolve later. For now, AD takes precedence.)

That covers all the “explored” systems. Note that I didn’t use the Zeb’s Guide data for the megacorp systems. I can always go back and add that but I don’t use those systems in my game. Finally for all the other systems, I had to randomly generate spectral types. To do that, I just generated a map with random values and copied the appropriate values over for the binary and single systems, only selecting single systems that were M stars or brown dwarfs. There are already way too many F, G, & K type stars in the region so I wanted a sprinkling of the other types. I also added the neutron star designation by hand for that system on the map.

Then it was time to actually run the program and make the map. It worked fine but looking at it I noticed a few problems with the jump distances. Several were too small by one and one was too large. Looking closer, I realized that while I was calculating the distances correctly, it was rounding wrong on display so I fixed that which corrected the distances displaying too small. That just left me with the one jump distance that was too large, now by two instead of one. This is where I discovered the typo on the original map for the distance between White Light and Madderly’s Star. The program was giving the correct distance but the original map was wrong.

I also decided that the symbols were a bit too small. The size I was using worked fine in the random map when there was the possibility of two systems in the same grid box with different Z-coordinates but for this map, where that couldn’t happen, I wanted the symbols larger. So I built in a scaling parameter in to the program allowing me to control the scale.

With those changes, we have the following initial color map of the Frontier.

You’ll notice a few things right away. First, there are no names on the map. Second, there are no nebulae. Right now the program doesn’t do those things.

Names will be relatively easy. I have the data, I just need to have the program add them to the image file. I need to research how to set fonts but otherwise, it’s straightforward.

Nebula are a whole different story. And may have to be added in by hand after the fact. Either way, that is going to be a complicated process and require some learning and work (and artistic ability) on my part.

If you look closer, you’ll notice that the positioning of the numbers along the jump routes are not optimal. It works for most of the jumps but for connections that are at an approximately 45 degree angle running from upper left to lower right, the number comes out too close (and sometimes overlapping) the line. I’ll need to fix that. That cleanup will probably be done by hand.

The program generates SVG files where each item is an object that can be manipulated allowing me to tweak things by hand. That was done intentionally as I knew the program wouldn’t be able to get everything perfect. That will be even more apparent once the names are added. I suspect that many of them will have to be moved to not overlap the jump routes.

The other minor thing is that the scaling factor I added wasn’t able to affect the size of the jets coming out of the neutron star. I need to either figure that out or, more likely, reduce the size back down for that spectral type and ignore the scaling. We’ll see how that goes.

Next Steps

The easy next steps are adding in the names of the systems and fixing the scaling on the neutron star.

After that, what I want to do is add in a way to handle all the various spectral types. The colors I picked don’t exactly represent the colors of the stars. They are close, but were chose for aesthetics rather than complete accuracy. Almost everything would be white with just some color tinting if we went for accuracy. I did that originally but didn’t like the result. I want to add code that will extrapolate between the given colors for the various spectral types’ number values to give various shades to the stars.

After that, I plan on tackling the nebula. I want to see if I can programmatically come up with something passable to start and then refine it.

The Code

If you want to play with the code, you can find it on my GitHub repository. It’s written in Python and right now isn’t the most user friendly. I need to add some in-line documentation as well as add features that will make it more usable directly from a command line instead of having to modify parameters in the code itself before running. That is another future project. But if you want to dive in and play with it as is, feel free.